The area of Groningen is most seismically active in the Netherlands. Earthquakes, triggered by gas extraction, cause not only physical damage but also social unrest. Various stakeholders need the earliest information after an earthquake occurs. Local authorities and emergency responders have set up PublicSonar to monitor the region for earthquake detection. By doing so, they ensure to never miss citizen signals for earthquakes and get the quickest alert.
Earthquake detection within 16 seconds with context, location & sentiment
During the night of 15 to 16 November 2021 at 01:46 an earthquake occurred in Garrelsweer, Groningen. Within the first seconds PublicSonar already received the first 4 messages, coming from local citizens. These messages not only provided information about the earthquake itself, but also about the location and sentiment of the citizens. One minute later, another 10 messages about the earthquake were recognised and signalled, making the magnitude of the impact clearer.
Within the same minute (01:47) of the first online messages and their validation, PublicSonar sent automated alerts by e-mail and SMS to local authorities and emergency services.
The automatic alerts and messages from PublicSonar were not only accurate, but also real-time and provided context about the impact of the earthquake. The alerts thus gave first responders and crisis managers an immediate head start in their actions. In comparison, the first ‘official’ public reports appeared only 19 minutes after the earthquake struck. Some popular news media did not make their first report until 53 minutes later.
A case like this clearly demonstrates the power of social media and other online messages. Especially in combination with technological innovation. PublicSonar’s AI solution supports organisations facing a crisis with the timely gathering of information. In order to act as quickly as possible and in the right way.
Open data: essential for managing crises
Every second at least 6,000 new tweets appear and in a day that is up to 500 million messages! The number of people sharing messages about crises and incidents has increased enormously in recent years. These messages are of great value to emergency response workers and crisis managers, for example: rapid (real-time) alerting of events, or the context provided by the messages in terms of sentiment, environment or location. This information also offers the possibility to react adequately. From the context, for example, it can become clear that (physical) resources need to be allocated.
How to utilise crowdsourced information for emergency response and crisis management?
During crises, incidents or disruptions of public safety, we see a sharp increase in the number of messages shared in a short period of time. This results in a large information stream, making it more challenging to extract the right context from the messages. Viewing and selecting messages ‘manually’ therefore becomes almost impossible. Let alone that useful conclusions or information can be extracted from them. As a result, there is a chance that crucial information will be missed because it is lost in the (too) large supply of information.
Intelligent and automatic filters are needed to extract the truly useful information from the large amount of data that emergency responders and crisis managers have to deal with. That is exactly what PublicSonar is unique in. Our algorithms collect data, filter it for relevance and recognise emotions. The technology automatically identifies incidents and alerts the right stakeholders and responders. Read more about our AI solution and discover the latest features.